公交车行程时间预测模型资料.doc

PAGE PAGE 1 公交车行程时间预测模型 摘要：随着城市道路交通面临着巨大的压力，交通问题已成为制约着城市经济发展的重要因素。公交车辆的按时乘坐成为了乘客最为关注的信息之一，而公交车行程时间预测这项技术应用对提高公交吸引力、提升乘客的满意度具有重要的意义，因此，本文以公交车运行时间预测为出发点，对其预测算法进行深入分析，以求探讨和研究更加精准的预测手段和方式。从技术的可行性和现实可行性的角度对公交到站时间预测技术进行了分析，总结现有预测算法的不同之处后，提出了将算法根据车辆运行的不同状态进行预测的方法，即为路段运行时间和车站服务时间两部分。本文在以公交车站调研基础之上，将影响预测结果的各种因素进行归类并且分布统计，提出了以支持向量机为基础、通过车辆实时车速而进行修正的路段运行时间预测模型，和以停车延误为基础进行的车站服务时间预测算法。本文的最后，选取了乌鲁木齐市个别公交车作为研究，利用调查数据对其车辆到站进行预测模型的验证，并和实际运行状况进行对比分析，验证了这些方法具有较高的正确性和可靠性。 关键词：智能公交；支持向量机；运行时间延误分析；预测结果分析 Research?on?the?Predicting?Model?of?Bus?Travel?Time Abstract: As the urban road traffic faced with tremendous pressure, traffic issues have become important factors of restricting the development of the urban economy. Bus ride as passengers were among the most concerned about the information in a timely manner, and bus travel time prediction technology applied to improve the public transport attractive, improve passenger satisfaction is of great significance, therefore, to forecast bus running time as the starting point of this article, its in-depth analysis of prediction algorithms in order to explore and study the more accurate means of predicting and ways. From the perspective of technical feasibility and practical feasibility of bus arrival time prediction technique for the analysis, after summing up the available prediction algorithms differ, proposed a algorithm based method for prediction of vehicle operation in different conditions, is the link run time and service to the two parts of the station. This article is based on the investigation of the bus station, factors influencing the predictions for the collation and distribution of statistics, presented based on support vector machine, Key words: Intelligent bus information system； SVM；Running time delay analysis；Prediction result analysis 1绪论 1．1 研究背景 伴随着社会经济的快速发展和城镇化进程加快，我市的机动车拥有量和城市道路交通量急剧增多；与之相适应，道路基础设施建设水平也是突飞猛进。但是，很明显后者的增长速度远远无法满足前者，同时我市的道路交通秩序管理上也存在一些问题，综合导致了道路交通的供需紧张，进而造成了城市中普遍存在的交通拥挤、事故频繁和环境污染等现象，交通问题已成为现代社会尤其是城市面临着的重